Friday, January 04, 2008

Sunshine duration vs Cloud Cover

One would expect a significant relationship between sunshine duration and total cloud cover. Greater cloud equals less sunshine, and less cloud equals greater sunshine.

And just to test this I looked at the relatioship between sunshine duration and total average monthly cloud cover at 3pm. Results were highly significant for at eh Giles weather station (t = -6.33, p < 0.001), as well as at Darwin (t = -4.77, p < 0.001) and at Tennant Creek (t = -3.03, p < 0.01).

So expectedly, the variable sunshine duration is a good measure of cloud cover throughout the whole day. But even better than that, we have a very extensive data set of total cloud cover (in eights) Australia wide.

So instead of having to look at the variable Sunshine duration, we can look at total cloud cover at eight different times of the day, too see what effect, if any, it has on recent warming trends.

And the results, will startle you....

3 comments:

Rafe said...

Don't keep us in suspense!

Jonathan Lowe said...

the data analysis is a big job, and due to increased work of late, it is slow going, but t will get there soon.

Anonymous said...

I have a mathematically sound proof that for a given instant:

1-P(S) =
1
Integral C * f(C) dC
0
where
P(S) = Probability for sunshine
(S = 0 or 1)
C = Cloud Cover
(0 ... 1)
f(C) = probability density
function of cloud cover

This formula just says that the average of "relative non sunshine hours" is equal to the "mean cloud cover".

To verify the formula one could - for example - take a sample of Cloud Cover readings at a given hour of the day at a given calendar month and compare it with the "relative non sunshine duration" over the same period.